Methods for X-ray imaging, X-ray imaging system and computer program product

Abstract

In X-ray imaging, an input image (9) and an input mask image (8) are generated based on the first and second detector data from a flat-panel X-ray detector (3). Noise reduction is then applied to generate a mask signal image (11) and a signal image (10). A subtraction image (13) is generated by calculating the difference between the signal image (10) and the mask signal image (11). The subtraction image (13) is scaled so that the number of pixels is increased in one direction. A result image (15) is generated by adding the scaled subtraction image (14) and a noise image (17, 19).

Description

[0001] The present invention relates to a method for X-ray imaging, wherein detector data are obtained from a flat-panel X-ray detector. The invention further relates to a data processing device for carrying out such a method, an X-ray imaging system comprising such a data processing device and a flat-panel X-ray detector, and corresponding computer software products.

[0002] In various X-ray-based imaging techniques, a mask image is subtracted from another X-ray image, also referred to as a live image or, in this disclosure, as an input image, of the same area. This subtraction highlights differences between the mask image and the live image and subtracts irrelevant parts of the X-ray image, such as anatomical background. Examples include digital subtraction angiography (DSA) and roadmap techniques. Depending on the specific technique, the mask image may be acquired without contrast agent and / or without a tool, whereas the live image may be acquired with contrast agent and / or the tool.In the resulting subtraction image, a vessel structure and / or the tool highlighted by the contrast agent is clearly visible, whereas disturbing or irrelevant image components, such as the image background, which are also present in the mask image, are not shown in the subtraction image.

[0003] X-ray flat panel detectors, for example, can be used to capture the mask image and the live image. These detectors have a large number of individual detectors or detector pixels arranged in a detector array along a first and a perpendicular second image direction. Various applications employ so-called zoom levels, in which only a portion, particularly a rectangular section, of the entire X-ray flat panel detector is used to generate the overall image. For display on a screen that has a higher resolution (in terms of a higher pixel count) than the used portion of the detector array, at least in one image direction, but usually in both, the subtracted image is upscaled to the target resolution of the screen. In other words, a digital zoom is performed. This process utilizes well-known interpolation methods.

[0004] The interpolation methods used can be based, for example, on bilinear or bicubic interpolation, on edge-adaptive interpolation, on superresolution approaches and / or on the use of artificial neural networks.

[0005] One disadvantage arises from the fact that X-ray imaging uses the lowest possible doses of X-rays, which leads to a comparatively high noise level. Upscaling and interpolating the subtraction image to the target resolution results in reduced image sharpness, particularly in the signal component of the subtraction image, and also alters the visual representation of the noise component.

[0006] One way to improve the resulting image quality would be to apply noise reduction algorithms to the subtracted image to reduce the noise level. However, since known noise reduction algorithms are not completely accurate, this can potentially create artifacts that may bother the user and also make it difficult or distort the interpretation of the displayed image.

[0007] Regarding the noise level, it should also be noted that the visual representation of the upscaled noise varies at different zoom levels. This leads to different image impressions at different zoom levels, which can further complicate or even distort the interpretation of the displayed image.

[0008] In the publication SM Rabiul Islam et al.: “Wavelet-based denoising algorithm of the ECG signal corrupted by WGN and Poisson noise,” 2012 International Symposium on Communications and Information Technologies (ISCIT), Gold Coast, QLD, Australia, 2012, pp. 165–168, a noise reduction algorithm based on a wavelet transform is described. In the publication M. Elad et al.: “Image Denoising Via Sparse and Redundant Representations Over Learned Dictionaries,” IEEE Transactions on Image Processing, vol. 15, no. 12, pp. 3736–3745, Dec. 2006, a noise reduction algorithm based on the non-local means approach is described. The publication J. Lehtinen et al.: “Noise2Noise: Learning Image Restoration without Clean Data”, arXiv:1803.04189 describes a method for noise reduction in images that is based on the use of artificial noise.Furthermore, noise reduction algorithms exist based on the principle of Singular Spectrum Analysis. This method uses Singular Spectrum Analysis to decompose a signal into its components. The noise within these components can then be identified and removed. Other noise reduction algorithms are based on convolutional artificial neural networks (CNNs). These CNNs are trained to remove noise from images or signals.

[0009] From the publication DE 10 2019 217 220 A1, a method for providing an output data set depending on an input data set is known, comprising the steps: - Determination of a noise-reduced signal data set from the input data set, - Determining a noise data set as the difference between the input data set and the signal data set, - Determining a modified noise data set from the noise data set and / or determining a modified signal data set from the signal data set, and - Determining an output data set by adding the modified noise data set to the signal data set or the modified signal data set, or by adding the noise data set to the modified signal data set.

[0010] From publication DE 10 2021 208 272 A1, a method for reducing noise and motion artifacts when generating a subtraction image is known. Several mask images of an object are acquired before the administration of a contrast agent. An image of the object is acquired after the contrast agent is administered. A first summed image is obtained by weighted summation of the mask images, whereby the individual weights of the mask images are automatically determined by an optimization procedure. The subtraction image is obtained by subtracting the summed image from the image.

[0011] From publication US 2019 / 0035058A1, a method for processing an X-ray image is known that includes a variance stabilization transformation. In this method, a transformation parameter of the variance stabilization transformation depends on a property of the X-ray image, which in turn depends on the X-ray imaging device and / or an acquisition parameter of the X-ray image. Noise-reduced data is generated from the variance-stabilized data, and a denoised X-ray image is then produced from this noise-reduced data using an inverse variance stabilization transformation.

[0012] From publication US 2017 / 0154413A1, a method for denoising a dynamic image is known. The dynamic image can represent a time series of snapshot images. The dynamic image is transformed into an aggregated image and a set of transformation domain images using a sparsifying transformation. The transformation domain images represent kinetic information of the dynamic images (i.e., differences between the snapshots), and the aggregated image represents static information (i.e., features and structures common to the snapshots). The transformation domain images are denoised. The denoised transformation domain images are recombined with the aggregated image using an inverse sparsifying transformation to produce a denoised dynamic image.

[0013] It is an object of the present invention to at least partially overcome the aforementioned disadvantages that result from upscaling the noise component.

[0014] This problem is solved by the subject matter of the independent claim. Advantageous further developments and preferred embodiments are the subject matter of the dependent claims.

[0015] The invention is based on the idea of ​​reducing noise by noise suppression in both the mask image and the input image before generating the subtraction image, generating the subtraction image and scaling it using interpolation, and then intentionally adding noise back to the scaled subtraction image.

[0016] According to one aspect of the invention, a method for X-ray imaging is described. First detector data is obtained from a flat-panel X-ray detector, and based on this first detector data, an input mask image is generated, representing a region of an object to be imaged in a first state. Second detector data is obtained from the same flat-panel X-ray detector, and based on this second detector data, an input image is generated, representing the region of the object to be imaged in a second state. By applying noise reduction, in particular a noise reduction algorithm, to the input mask image, a mask signal image is generated. By applying noise reduction, in particular a noise reduction algorithm or a further noise reduction algorithm, to the input image, a signal image is generated.A subtraction image is generated by taking a difference between the signal image and the mask signal image, or between the signal image and an image dependent on the mask signal image, or between an image dependent on the signal image and the mask signal image, or between the image dependent on the signal image and the image dependent on the mask signal image.

[0017] A scaled subtraction image is generated by scaling the subtraction image using an interpolation method, in particular by upscaling it, such that at least in one first image direction, the number of pixels in the scaled subtraction image is greater than the number of pixels in the subtraction image. A result image is generated by adding the scaled subtraction image and a noise image.

[0018] The method according to the invention can be purely computer-implemented. Unless otherwise specified, all steps of the computer-implemented method can be performed by a data processing device comprising at least one processing unit. In particular, the at least one processing unit is configured or adapted to perform the steps of the computer-implemented method. For this purpose, the at least one processing unit can, for example, store a computer program containing instructions which, when executed by the at least one processing unit, cause the at least one processing unit to execute the computer-implemented method.

[0019] For each embodiment of the computer-implemented method, corresponding embodiments of the method according to the invention result which are not purely computer-implemented, by including corresponding method steps according to which the first detector data and / or the second detector data are generated by means of the X-ray flat panel detector.

[0020] Unless explicitly stated otherwise below, an image always refers to an image taken in the local area.

[0021] The input mask image and the input image have, in particular, the same number of pixels and the same aspect ratio. Accordingly, the subtraction image, for example, also has the same number of pixels as the input mask image and the input image before scaling.

[0022] The resulting image, or an image dependent on the resulting image, can be displayed on a screen, for example.

[0023] The input image and the input mask image are, in particular, X-ray projection images. To generate the respective detector data, X-rays are emitted from an X-ray source in the direction of the area of ​​the object to be imaged, and the detector data are generated by the X-ray flat panel detector based on the portions of the X-rays passing through the object.

[0024] The first state exists during a first period, during which the first detector data is generated, and the second state exists during a second period, during which the second detector data is generated. Specifically, the second period lies after the first.

[0025] The first and second states can differ, for example, by the presence of a contrast agent in the area to be imaged, which in particular contains one or more blood vessels. The first state corresponds, for example, to a state in which no contrast agent is present in the area to be imaged, and the second state corresponds to a state in which contrast agent is present in the area to be imaged. The administration of the contrast agent can, for example, take place between the first and second periods, but as such is generally not part of the method according to the invention. The subtraction image then shows exclusively the area in which the contrast agent is present, i.e., the one or more blood vessels, or the area in which the contrast agent is present is particularly emphasized in the subtraction image and is therefore clearly visible.

[0026] The first state and the second state can differ, for example, additionally or alternatively, by the presence of a tool in the area to be imaged. The first state corresponds, for example, to a state in which the tool is not located in the area to be imaged or a defined sub-area thereof, and the second state corresponds to a state in which the tool is located in the area to be imaged or the sub-area thereof. The introduction of the tool can, for example, occur between the first and second periods, but as such is generally not part of the method according to the invention. The subtraction image then shows exclusively the tool, or the tool is particularly emphasized in the subtraction image and is therefore clearly recognizable.The resulting image can also be superimposed on at least one other image when displayed on the monitor, for example, an image that only shows the vessels into which the tool is inserted or highlights them. The tool could be, for example, a vascular catheter and / or a vascular prosthesis, a stent, a guide wire, or the like.

[0027] The initial detector data can correspond to a single acquisition interval, also known as a frame, resulting in exactly one X-ray projection image, namely the input mask image. However, it is also possible for the initial detector data to correspond to multiple frames. In this case, several X-ray projection images can be generated based on the initial detector data, one for each frame. The input mask image can then correspond to one of the X-ray projection images or be generated by averaging or weighted averaging the X-ray projection images.

[0028] It is also possible that, when selecting the X-ray projection image or performing weighted averaging, a similarity measure is calculated between the X-ray projection image and the input image. Depending on the similarity measure, one of the X-ray projection images is then selected as the input mask image, or corresponding weighting factors for weighted averaging are calculated based on the similarity measure. This ensures that the content of the input image, which is also present in the input mask image and is to be removed, particularly during the generation of the subtraction image, matches that of the input mask image as closely as possible. The similarity measure can be calculated, for example, as an L2 distance, or more complex measures can be used, such as those employing segmentation. The second detector data can also correspond to one or more frames and consequently to one or more X-ray projection images. The input image is selected, for example, from the X-ray projection images.

[0029] The noise reduction algorithm, or any subsequent noise reduction algorithm, can be a known noise reduction algorithm. For example, it can be implemented according to the publications on noise reduction algorithms cited at the beginning. The interpolation method can also be based on one of the aforementioned interpolation methods.

[0030] Noise reduction results in a lower noise component in the signal image than in the input image, ideally eliminating it entirely, and a lower noise component in the masked signal image than in the input mask image, ideally eliminating it entirely. For example, the signal-to-noise ratio of the signal image is higher than the signal-to-noise ratio of the input image, and the signal-to-noise ratio of the masked signal image is higher than the signal-to-noise ratio of the input mask image.

[0031] In conventional approaches where noise reduction is not performed before generating and scaling the subtraction image, the visual representation of the upscaled noise component depends on the respective zoom level. This dependency is avoided according to the invention. In particular, the noise image can be generated in such a way that it results in the same or substantially the same visual representation for all zoom levels. Even if noise reduction were performed in conventional methods after generating and before scaling the subtraction image, or after scaling the subtraction image, this would potentially lead to undesirable artifacts. This, too, is avoided according to the invention.

[0032] Furthermore, it has been shown that adding the noise image to the scaled subtraction image unexpectedly improves the visual impression when displaying the result image, in that a trained viewer can better perceive anatomical structures and / or other content depicted in the result image. This also improves the independence of the visual impression when displaying the result image from the zoom level.

[0033] The noise image is generated in such a way that the respective number of pixels in the first and second image directions of the noise image is equal to the respective number of pixels in the first and second image directions of the scaled signal image. The individual pixel values ​​can therefore be added to generate the resulting image.

[0034] In some embodiments, the noise pattern can be generated independently of the first detector data and / or the second detector data. For example, white noise can be used to generate the noise pattern. However, it is also possible to generate the noise pattern depending on the first detector data and / or the second detector data. This has the advantage that the noise characteristics of the X-ray flat panel detector are essentially maintained.

[0035] According to at least one embodiment, an initial noise image is generated by calculating the difference between the input image and the signal image. The noise image is generated by extending a spatial frequency spectrum of the initial noise image or of an image dependent on the initial noise image, such that a maximum spatial noise frequency is increased corresponding to the first image direction.

[0036] Through noise reduction and subtraction (i.e., subtracting the input image from the signal image, or vice versa), the input image is split into the signal image and the initial noise image. Ideally, the signal image contains no noise components whatsoever, whereas the initial noise image would contain only noise components, specifically all noise components of the input image. However, depending on the input image and the noise reduction algorithm used, this ideal case is generally not achieved. Nevertheless, noise components dominate in the initial noise image, whereas they play a subordinate role in the signal image.

[0037] The initial noise image is a discrete digital image. Its frequency spectrum can therefore be calculated using a discrete Fourier transform, such as a Fast Fourier Transform (FTT). Since the initial noise image is two-dimensional, the frequency spectrum is also two-dimensional. In certain embodiments, the frequency spectrum is extended in the first image direction. However, in some embodiments, the frequency spectrum can be extended in both the first and second image directions, thus increasing the maximum noise frequency in both directions.

[0038] The extension of the frequency spectrum can be performed in various ways. For example, the extension can be carried out in the frequency domain. For this, the frequency spectrum of the initial noise image is calculated, in particular using the aforementioned discrete Fourier transform. Due to the discrete nature of the initial noise image, there exists a maximum spatial noise frequency of the frequency spectrum corresponding to the first image direction and another maximum noise frequency corresponding to the second image direction. The maximum noise frequency is given by the inverted pixel size of the initial noise image in the first image direction, and the other maximum noise frequency by the inverted pixel size of the initial noise image in the second image direction.

[0039] If the initial noise image were upscaled in the conventional manner, so that the number and size of pixels in the upscaled initial noise image corresponded to the scaled subtraction image, the corresponding amplitudes of the frequency components would be zero at high frequencies in the scaled initial noise image. This would lead to an inconsistent noise impression or an inconsistent representation of the noise component if the scaled initial noise image were added to and displayed with the scaled subtraction image. Therefore, in corresponding embodiments, the frequency spectrum is extended such that the maximum noise frequency of the initial noise image is increased, at least along the first image direction.

[0040] If necessary, appropriate normalizations are also carried out to adapt the pixel size and number of pixels of the noise image to those of the scaled subtraction image.

[0041] The extension of the initial noise image's frequency spectrum can be understood, in particular, as supplementing it with frequency components corresponding to the first image direction, thus increasing the maximum spatial noise frequency. For example, the added frequency components can be determined based on white noise. Alternatively, the initial noise image can be scaled analogously to the subtraction image, and a modulation transfer function (MTF) can be calculated as a function of the frequencies f1 and f2 corresponding to the first and second image directions, respectively. MTF(f1,f2)=1−A′(f1,f2)A(f1,f2), where A(f1,f2) represents the corresponding amplitude in the frequency domain before scaling and A'(f1,f2) the corresponding amplitude in the frequency domain after scaling. The MTF can be convolved with a white noise spectrum and the result added to A'(f1,f2). Transforming the sum back into spatial space by inverse Fourier transform yields, for example, the noise pattern.

[0042] In other embodiments, however, the frequency spectrum can also be extended directly in spatial space. For this purpose, the initial noise image can be scaled analogously to the subtraction image, and amplitude values ​​can be statistically assigned to those resulting pixels where no spatial amplitude values ​​are present in the scaled initial noise image, according to a predefined variance. This variance can correspond to a variance of the original initial noise image.

[0043] The initial noise image can also be decomposed into multiple frequency bands in a multi-scalar approach, for example, by applying the principle of Laplace pyramids. The resulting frequency-band-specific initial noise images are then scaled like the scaled subtraction image. A variance can be calculated for each frequency-band-specific initial noise image, and the unoccupied pixels in the frequency-band-specific scaled initial noise images can be filled with statistically distributed amplitude values ​​according to the variances thus determined. The noise image can then be generated by recomposing the augmented frequency-band-specific scaled initial noise images, particularly according to the principle of Laplace pyramids.

[0044] This is particularly advantageous if the noise in the initial noise image is colored, for example, due to the MTF of the X-ray flat panel detector. Since the variance of Poisson noise is amplitude-dependent, it is especially advantageous to apply a variance-stabilizing transformation beforehand, as described below for embodiments. If the Poisson noise is superimposed with electrical noise, the coloration, in addition to the variance, is also a function of the level. In this case, a variance- and covariance-stabilizing transformation would be even more advantageous.

[0045] Thus, in appropriate embodiments, even with different input image and input mask image sizes, the same noise characteristics can always be achieved in the resulting image by extending the frequency spectrum to the same maximum noise frequency. Ideally, a user cannot therefore draw any conclusions from the result image about the size of the input image or input mask image, or the corresponding proportion or section of the X-ray flat panel detector array used. In other words, the visual representation of the noise component in the result image does not differ for different input image or input mask image sizes. This is particularly advantageous because a consistent noise appearance is important for the reliable interpretation of the result image.

[0046] The above explanations concern the extension of the spatial frequency spectrum of the initial noise image. In such embodiments, the noise image is generated independently of the first detector data or the input mask image. This can be particularly advantageous if a significantly higher X-ray dose per frame is used to generate the first detector data than to generate the second detector data. In this case, the noise component in the input mask image is, for example, significantly lower than in the input image and can potentially be neglected.

[0047] The noise image can also be generated depending on the input image and the input mask image. For example, the image dependent on the initial noise image can depend on both the initial noise image and the input mask image. The above explanations can be applied analogously to embodiments in which, instead of extending the spatial frequency spectrum of the image dependent on the initial noise image, the spatial frequency spectrum of the image dependent on the initial noise image is extended.

[0048] According to at least one embodiment, a mask noise image is generated by calculating the difference between the input mask image and the mask signal image. The image dependent on the initial noise image is generated by adding or weighting the initial noise image and the mask noise image. The noise image is generated by extending the spatial frequency spectrum of the image dependent on the initial noise image, such that the maximum spatial noise frequency is increased according to the first image direction.

[0049] This further increases the consistency of the noise characteristics in the resulting image.

[0050] According to at least one embodiment, a first raw image is generated based on the first detector data by means of the at least one computing unit, and the input mask image is generated by applying a variance-stabilizing transformation, in particular an Anscombe transformation, to the first raw image.

[0051] According to at least one embodiment, a second raw image is generated based on the second detector data by means of the at least one computing unit, and the input image is generated by applying a variance-stabilizing transformation, in particular an Anscombe transformation, to the second raw image.

[0052] The variance-stabilizing transformation ensures that the variance of the noise in the input mask image or the input image is at least approximately independent of the level, i.e., the local mean of the noise. In the raw images, for example, the variance of the noise is approximately proportional to the mean, since the noise roughly follows a Poisson distribution.

[0053] The variance-stabilizing transformation ultimately makes it possible to ensure that the distribution of the noise component in the resulting image is independent of the brightness of the displayed image components.

[0054] In some embodiments, the variance-stabilizing transformation is a variance- and covariance-stabilizing transformation.

[0055] As mentioned above, this allows the dependence of the noise's color on the noise level to be compensated for.

[0056] According to at least one embodiment, the image dependent on the signal image is generated by performing at least one image processing step.

[0057] According to at least one embodiment, the at least one image processing step includes an inverse variance-stabilizing transformation, in particular the inverse transformation to the variance-stabilizing transformation that is applied to the second raw image to generate the input image.

[0058] This ensures that the effects of the variance-stabilizing transformation only occur during the processing of the initial noise image, and not during the processing of the signal image. This avoids artifacts that would otherwise occur in the signal path.

[0059] According to at least one embodiment, the at least one image processing step includes at least one grayscale transformation.

[0060] In this way, contrasts can be improved in the signal path. The grayscale transformation can, for example, include a logarithmic grayscale transformation. This involves calculating the logarithm of the corresponding pixel values ​​in the signal path to achieve a homogeneous brightness distribution in the resulting image, in accordance with the exponential nature of the attenuation law for X-ray radiation.

[0061] According to at least one embodiment, the at least one image processing step includes the application of at least one frequency filter, for example at least one bandpass filter and / or at least one multiscalar filter.

[0062] According to at least one embodiment, the image dependent on the mask signal image is generated by performing at least one further image processing step.

[0063] According to at least one embodiment, the at least one further image processing step includes an inverse variance-stabilizing transformation, in particular the inverse transformation to the variance-stabilizing transformation that is applied to the first raw image to generate the input mask image.

[0064] This ensures that the effects of the variance-stabilizing transformation only occur during the processing of the mask noise image, and not during the processing of the mask signal image. This avoids artifacts that would otherwise occur in the signal path.

[0065] According to at least one embodiment, the at least one further image processing step includes at least one further grayscale transformation.

[0066] This allows for improved contrast in the signal path. Further grayscale transformation can, for example, include a logarithmic grayscale transformation.

[0067] According to at least one embodiment, the at least one further image processing step includes the application of at least one frequency filter, for example at least one bandpass filter and / or at least one multiscalar filter.

[0068] According to at least one embodiment, the number of pixels in the scaled subtraction image in the first image direction corresponds to the number of pixels in the screen in the first image direction.

[0069] According to at least one embodiment, the increased maximum spatial noise frequency corresponds to an inverted pixel size of the screen in the first image direction.

[0070] According to at least one embodiment, the first detector data include respective pixel values ​​for detector pixels of a predetermined section of the detector array of the X-ray flat panel detector, in particular all detector pixels of the predetermined section and only the detector pixels of the predetermined section.

[0071] The input mask image is therefore an input mask image corresponding to a predefined zoom level that corresponds to the predefined section.

[0072] According to at least one embodiment, the second detector data includes respective pixel values ​​for detector pixels of the specified section of the detector array of the X-ray flat panel detector, in particular all detector pixels of the specified section and only the detector pixels of the specified section.

[0073] The input image is therefore an input image corresponding to the specified zoom level, which corresponds to the specified section.

[0074] According to at least one embodiment, the subtraction image is scaled using the interpolation method, such that in the second image direction, which is in particular perpendicular to the first image direction, the number of pixels in the scaled subtraction image is greater than the number of pixels in the subtraction image.

[0075] According to at least one embodiment, to generate the noise image, the spatial frequency spectrum of the initial noise image or of the image dependent on the initial noise image is extended in such a way that a further maximum spatial noise frequency is increased according to the second image direction.

[0076] The above statements regarding the increase of the maximum spatial noise frequency according to the first image direction can be applied analogously to the increase of the further maximum spatial noise frequency according to the second image direction.

[0077] According to at least one embodiment, X-rays are emitted by means of an X-ray source in the direction of the area of ​​the object to be imaged while the object is in the first state, and the first detector data are generated by the X-ray flat panel detector based on the portions of the X-rays passing through the area of ​​the object to be imaged. Further X-rays are emitted by means of the X-ray source in the direction of the area of ​​the object to be imaged while the object is in the second state, and the second detector data are generated by means of the X-ray flat panel detector based on the portions of the further X-rays passing through the area of ​​the object to be imaged.

[0078] According to a further aspect of the invention, a data processing device is provided which has at least one computing unit. The at least one computing unit is adapted to carry out a method for X-ray imaging according to the invention, in particular a computer-implemented method according to the invention.

[0079] A computing unit can be understood, in particular, as a data processing device containing a processing circuit. The computing unit can therefore process data to perform arithmetic operations. This may also include operations to perform indexed access to a data structure, such as a lookup table (LUT).

[0080] The computing unit may, in particular, contain one or more computers, one or more microcontrollers, and / or one or more integrated circuits, for example, one or more application-specific integrated circuits (ASICs), one or more field-programmable gate arrays (FPGAs), and / or one or more systems on a chip (SoCs). The computing unit may also contain one or more processors, for example, one or more microprocessors, one or more central processing units (CPUs), one or more graphics processing units (GPUs), and / or one or more signal processors, in particular one or more digital signal processors (DSPs). The computing unit may also include a physical or virtual array of computers or other units of the aforementioned type.

[0081] In various embodiments, the computing unit includes one or more hardware and / or software interfaces and / or one or more storage units.

[0082] A storage unit can be volatile data storage, for example as dynamic random access memory (DRAM) or static random access memory (SRAM), or as non-volatile data storage, for example as read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), flash memory or flash EEPROM, ferroelectric random access memory (FRAM), or magnetoresistive random access memory.It can be designed as MRAM (magnetoresistive random access memory) or as phase-change random access memory, PCRAM (phase-change random access memory).

[0083] According to a further aspect of the invention, an X-ray imaging system is provided, comprising an X-ray source and a flat-panel X-ray detector. The X-ray imaging system includes a data processing device according to the invention, wherein the at least one computing unit is configured to perform an X-ray imaging method according to the invention and, for this purpose, to control the X-ray source to generate the X-ray radiation and the subsequent X-ray radiation, and to obtain the first detector data and the second detector data from the X-ray detector.

[0084] According to at least one embodiment of the X-ray imaging system according to the invention, it has a screen and the at least one computing unit is configured to display the result image or an image dependent on the result image on the screen.

[0085] Further embodiments of the X-ray imaging system according to the invention follow directly from the various configurations of the method according to the invention, and vice versa. In particular, individual features and corresponding explanations as well as advantages relating to the various configurations of the method according to the invention can be transferred analogously to corresponding configurations of the X-ray imaging system according to the invention. In particular, the X-ray imaging system according to the invention is configured or programmed to carry out a method according to the invention. In particular, the X-ray imaging system according to the invention carries out the method according to the invention.

[0086] According to a further aspect of the invention, a first computer program comprising first instructions is specified. When the first instructions are executed by a data processing device, in particular a data processing device according to the invention, the first instructions cause the data processing device to execute a method for X-ray imaging according to the invention, in particular a computer-implemented method according to the invention.

[0087] The initial instructions can be provided as program code. This program code can be, for example, binary code or assembly language, and / or source code in a programming language such as C, and / or a program script such as Python.

[0088] According to a further aspect of the invention, a second computer program with second commands is specified. When the second commands are executed by an X-ray imaging system according to the invention, in particular by the data processing device of the X-ray imaging system, the second commands cause the X-ray imaging system to perform an X-ray imaging method according to the invention.

[0089] The second set of instructions can be provided as program code. This program code can be, for example, binary code or assembly language, and / or source code in a programming language such as C, and / or a program script such as Python.

[0090] According to another aspect, a computer-readable storage medium is specified which stores a first computer program according to the invention and / or a second computer program according to the invention.

[0091] The first computer program, the second computer program, and the computer-readable storage medium can each be referred to as computer program products containing the first and second instructions, respectively.

[0092] Further features and combinations of features of the invention will become apparent from the figures and their description, as well as from the claims. In particular, further embodiments of the invention need not necessarily include all features of any one of the claims. Further embodiments of the invention may have features or combinations of features not mentioned in the claims.

[0093] The invention is explained in more detail below with reference to specific embodiments and associated schematic drawings. In the figures, identical or functionally equivalent elements may be designated with the same reference numerals. The description of identical or functionally equivalent elements is not necessarily repeated with respect to different figures.

[0094] The figures show: Fig. 1 a schematic representation of an exemplary embodiment of an X-ray imaging system according to the invention; Fig. 2 a schematic flowchart of an exemplary embodiment of a method according to the invention for X-ray imaging; Fig. 3 a schematic flowchart of a further exemplary embodiment of a method according to the invention for X-ray imaging; Fig. 4 a schematic flowchart of a further exemplary embodiment of a method according to the invention for X-ray imaging; and Fig. 5 a schematic representation of a spatial frequency spectrum of an initial noise image or a noise image according to a further exemplary embodiment of an inventive method for X-ray imaging.

[0095] In Fig. Figure 1 schematically represents an exemplary embodiment of an X-ray imaging system 1 according to the invention, comprising an X-ray source 4 and a flat X-ray detector 3, as well as at least one computing unit, which for the sake of simplicity is shown in Fig. Figure 1 shows a single computing unit 2. An object 5, in particular a body part of a patient, can be placed in a beam path of the X-ray radiation emitted by the X-ray source 4. The portions of the X-ray radiation passing through the object 5 strike a detector array of the flat X-ray detector 3, and this detector generates detector data based on the detected portions of the X-ray radiation, in particular using a predefined section of the detector array, and provides this data to the computing unit 2.

[0096] The computing unit 2 can produce a result image 15 (see) by carrying out a method for X-ray imaging according to the invention. Fig. 2) generate and, if necessary, control a screen (not shown) of the X-ray imaging system 1 to display the result image 15 or a dependent image.

[0097] In Fig. Figure 2 is a schematic flowchart of an exemplary embodiment of a method for X-ray imaging according to the invention.

[0098] The processing unit 2 receives initial detector data from the X-ray flat panel detector 3 and, based on this initial detector data, generates an input mask image 8 that represents a region of the object 5 to be imaged, for example, a vascular tree, in a first state, for example, without contrast agent. Subsequently, for example, after the administration of contrast agent, the processing unit 2 receives second detector data from the X-ray flat panel detector 3 and, based on this second detector data, generates an input image 9, also referred to as a live image, that represents the region of the object 5 to be imaged in a second state, for example, with the contrast agent.

[0099] By applying noise suppression to the input mask image 8, the processing unit 2 generates a mask signal image 11, and by applying noise suppression to the input image 9, a signal image 10.

[0100] By calculating the difference between signal image 10 and mask signal image 10, or images dependent on each of these, the processing unit 2 generates a subtraction image 13. The processing unit 2 then generates a scaled subtraction image 14 by scaling subtraction image 13 using an interpolation method, such that in one image direction, the number of pixels in the scaled subtraction image 14 is greater than the number of pixels in subtraction image 13. The processing unit 2 generates the result image 15 by adding the scaled subtraction image 14 and a noise image 19. The noise image 19 can, for example, be an artificial noise image based on white noise.

[0101] In Fig. Figure 3 shows a schematic flowchart of a further exemplary embodiment of an X-ray imaging method according to the invention, which is based on the Fig. 2 described embodiment is based on.

[0102] Instead of noise image 19, a noise image 17 is added to the scaled signal image 14, which depends on the input image 9. An initial noise image 12 is generated by taking the difference between the input image 9 and the signal image 10. The noise image 17 is generated by a spatial frequency spectrum 6 (see Fig. 5) of the initial noise image 12 is extended so that a maximum spatial noise frequency is increased according to the first image direction.

[0103] In Fig. Figure 4 shows a schematic flowchart of a further exemplary embodiment of an X-ray imaging method according to the invention, which is based on the Fig. 2 described embodiment is based on.

[0104] Instead of noise image 19, a noise image 17 is added to the scaled signal image 14, which depends on the input mask image 8 and the input image 9. A mask noise image 18 is generated by taking the difference between the input mask image 8 and the mask signal image 11. The initial noise image 12 is generated by taking the difference between the input image 9 and the signal image 10. An image 16, which depends on the initial noise image 12 and the mask noise image 18, is generated by adding or weighting the initial noise image 12 and the mask noise image 18. The noise image 17 is generated by extending the spatial frequency spectrum 6 of image 16 so that the maximum spatial noise frequency is increased according to the first image direction.

[0105] In some embodiments, the input mask image 8 is generated by creating a first raw image based on the first detector data and applying a variance-stabilizing transformation to this raw image. A corresponding inverse variance-stabilizing transformation can then be applied to the mask signal image 11 before generating the subtraction image 13. In some embodiments, the input image 9 is generated by creating a second raw image based on the second detector data and applying a variance-stabilizing transformation to this raw image. A corresponding inverse variance-stabilizing transformation can then be applied to the signal image 10 before generating the subtraction image 13.

[0106] In some embodiments, one or more grayscale transformations, for example using one or more lookup tables, are performed on the mask signal image 11 and / or filters are applied for contrast processing before the subtraction image 13 is generated. In some embodiments, one or more grayscale transformations, for example using one or more lookup tables, are performed on the signal image 10 and / or filters are applied for contrast processing before the subtraction image 13 is generated.

[0107] In Fig. Figure 5 shows an exemplary way to supplement the frequency spectrum, with a corresponding frequency spectrum 6 of the initial noise image 12 shown along the first image direction. Due to the discrete nature of the initial noise image 12, the frequency spectrum 6 exhibits a maximum spatial noise frequency, which limits the frequency space.

[0108] To extend or supplement the frequency spectrum, the initial noise image 12 can, for example, first be scaled according to the subtraction image 13, so that in the first image direction the number of pixels of the scaled initial noise image corresponds to the number of pixels of the scaled subtraction image 14. The scaled initial noise image can then be transformed into the frequency domain by applying a discrete two-dimensional FFT.

[0109] The unscaled initial noise image 12 can also be transformed into the frequency domain. The amplitudes of the resulting frequency spectra of the initial noise image 12 and the scaled initial noise image can be used to determine a modulation transfer function, MTF, according to the equation MTF(f1,f2)=1−A′(f1,f2)A(f1,f2), to determine where A(f1,f2) represents the corresponding amplitude in the frequency space before scaling and A' (f1,f2) represents the corresponding amplitude in the frequency space after scaling.

[0110] The MTF can be convolved with an artificial noise image, for example, an artificial noise image representing two-dimensional white noise. The result of the convolution corresponds to additional frequency components 7, which can then be added to the frequency spectrum 6 of the scaled initial noise image. The result can then be transformed back into space space by an inverse FFT to obtain the noise image 17.

[0111] In alternative embodiments, the initial noise image 12 can be decomposed into sub-images corresponding to different frequency bands, for example by applying the Gaussian-Laplace pyramid method or the multiscale analysis of wavelet theory. The variances of the noise in the individual sub-images can then be determined, and artificial noise with the same variance but an increased, predefined resolution can be inserted into the individual sub-images. The noise image 17 can then be reconstructed by combining the added sub-images. In this way, the FFT can be avoided, thus reducing the computational effort.

[0112] For use cases or application situations that may arise during the procedure and are not explicitly described here, it may be provided that, according to the procedure, an error message and / or a request for user feedback is issued and / or a default setting and / or a predetermined initial state is set.

[0113] As described, particularly with reference to the figures, the invention can at least partially overcome disadvantages resulting from scaling up the subtraction image.

[0114] In various embodiments, this is made possible in particular by the separate processing of the noise and signal components. The signal component is processed using interpolation, for example, edge-adaptive interpolation, while the noise component of the image is modified so that it contains higher noise frequencies than the original image. In preferred embodiments, the noise component can be modified as if the pixel size did not change during scaling.

[0115] Regardless of the grammatical gender of any particular term used in the present disclosure, persons of male, female or other gender identities are included.

Claims

[1] A method for X-ray imaging, wherein - first detector data are obtained from an X-ray flat panel detector (3) and, based on the first detector data, an input mask image (8) is generated which represents a region of an object (5) to be imaged in a first state; - second detector data are obtained from the flat X-ray detector (3) and, based on the second detector data, an input image (9) is generated which represents the region of the object (5) to be imaged in a second state; - a mask signal image (11) is generated by applying noise suppression to the input mask image (8); - a signal image (10) is generated by applying noise suppression to the input image (9); - a subtraction image (13) is generated by forming the difference between the signal image (10) or an image dependent on the signal image (10) and the mask signal image (11) or an image dependent on the mask signal image (11); - a scaled subtraction image (14) is generated by scaling the subtraction image (13) using an interpolation method, so that at least in a first image direction a pixel number of the scaled subtraction image (14) is greater than a pixel number of the subtraction image (13); and - a result image (15) is generated by adding the scaled subtraction image (14) and a noise image (17, 19). [2] The method according to claim 1, wherein - an initial noise image (12) is generated by forming the difference between the input image (9) and the signal image (10); and - the noise image (17) is generated by expanding a spatial frequency spectrum (6) of the initial noise image (12) or of an image (16) dependent on the initial noise image (12) such that a maximum spatial noise frequency is increased in accordance with the first image direction. [3] Method according to claim 2, wherein - a mask noise image (18) is generated by forming the difference between the input mask image (8) and the mask signal image (11); - the image (16) dependent on the initial noise image (12) is generated by addition or weighted addition of the initial noise image (12) and the mask noise image (18); and - the noise image (17) is generated by expanding the spatial frequency spectrum (6) of the image (16) dependent on the initial noise image (12) so that the maximum spatial noise frequency is increased according to the first image direction. [4] Method according to one of the preceding claims, wherein a raw image is generated based on the second detector data and the input image (9) is generated by applying a variance-stabilizing transformation to the raw image. [5] The method of claim 4, wherein the variance stabilizing transformation includes an Anscombe transformation. [6] Method according to one of claims 4 or 5, wherein - the image dependent on the signal image (10) is generated by performing at least one image processing step, wherein the at least one image processing step includes an inverse variance-stabilizing transformation; and / or - the image dependent on the mask signal image (11) is generated by performing at least one further image processing step, wherein the at least one further image processing step includes an inverse variance-stabilizing transformation. [7] Method according to claim 6, wherein - the at least one image processing step includes at least one gray value transformation and / or an application of at least one frequency filter; and / or - the at least one further image processing step includes at least one further gray value transformation and / or an application of at least one further frequency filter. [8] Method according to claim 7, wherein - the at least one gray value transformation includes a logarithmic gray value transformation; and / or - the at least one further gray value transformation includes a logarithmic gray value transformation. [9] Method according to one of the preceding claims, wherein the result image (15) or an image dependent on the result image (15) is displayed on a screen. [10] The method according to claim 9, wherein the number of pixels of the scaled subtraction image (14) in the first image direction corresponds to a number of pixels of the screen in the first image direction. [11] Method according to one of the preceding claims, wherein the subtraction image (13) is scaled using the interpolation method so that in a second image direction a pixel number of the scaled subtraction image (14) is greater than a pixel number of the subtraction image (13). [12] Method according to one of the preceding claims, wherein - X-ray radiation is emitted by an X-ray source (4) in the direction of the area of ​​the object (5) to be imaged while the object (5) is in the first state, and the first detector data are generated by the flat X-ray detector (3) based on portions of the X-ray radiation passing through the area of ​​the object (5) to be imaged; and - by means of the X-ray source (4) further X-ray radiation is emitted in the direction of the region of the object (5) to be imaged, while the object (5) is in the second state, and the second detector data are generated by means of the X-ray flat detector (3) based on portions of the further X-ray radiation passing through the region of the object (5) to be imaged. [13] Data processing device comprising at least one computing unit (2) adapted to carry out a method according to one of claims 1 to 11. [14] X-ray imaging system (1) comprising an X-ray source (4), an X-ray flat detector (3), and a data processing device according to claim 13, wherein the at least one computing unit (2) is configured to carry out a method according to claim 12 and to control the X-ray source (4) to generate the X-ray radiation and the further X-ray radiation and to receive the first detector data and the second detector data from the X-ray detector. [15] Computer program product comprising - first instructions which, when executed by a data processing device, cause the data processing device to execute a method according to one of claims 1 to 11; or - second instructions which, when executed by an X-ray imaging system (1) according to claim 14, cause the X-ray imaging system (1) to carry out a method according to any one of claims 1 to 12.

Images